Disagreement between parents and children about a child's symptoms, termed parent/child informant discrepancies (PCIDs), is one of the most robust findings in the child clinical literature (Achenbach, McConaughy, & Howell, 1987; De Los Reyes, 2012). However, interpretation of these discrepancies is notoriously difficult (Smith, 2007). Despite advances in our understanding of how child and family characteristics relate to discrepancies (De Los Reyes et al., 2013), little is known of how PCIDs relate to treatment outcomes. Prior theoretical (Ferdinand, van der Ende, & Verhulst, 2004) and empirical (Kolko & Kazdin, 1993) work suggests greater discrepancies are associated with a host of factors thought to impede treatment outcomes (e.g., family functioning, willingness to discuss negative emotion), suggesting PCIDs likely represent important, easily collected, indicators for treatment prognosis. Consistent with this theory, preliminary evidence suggests higher agreement between parents and children regarding symptoms at the start of treatment, and increased concordance over the course of treatment, predict better outcomes (rated by an independent evaluator) for psychosocial anxiety treatment (De Los Reyes, Alfano, & Beidel, 2010; Panichelli-Mindel, Flannery-Schroeder, Kendall, & Angelosante 2005). However, to date, no studies have examined which patterns of PCIDs prior to (e.g., parent high/child low vs. child high/parent low) and throughout treatment (e.g., discrepancies that remain high or those that lessen over time) predict worse outcomes. Using data from the largest treatment trial of pediatric anxiety to date, the Child/Adolescent Anxiety Multimodal Study (CAMS; Walkup et al., 2008; Ginsburg et al 2011), the proposed study will apply the latest statistical advancements in the study of informant discrepancies (i.e., polynomial regression coefficients; Laird & De Los Reyes, 2013), as well as latent class growth modeling to examine how PCIDs prior to and throughout treatment predict treatment outcomes. Additionally, this study will also address how these relationships differ by treatment modality (psychosocial versus pharmacotherapy). Furthering this research is critical to understanding how PCIDs can be used to identify youth at risk for treatment failure or those in need of adjunctive treatments (e.g., family therapy). By addressing these questions, this proposal addresses NIMH Strategic Objective 3.2, which has the goal of expanding our understanding of clinical outcomes to understand individual variation in treatment response. Results from this study will greatly advance our understanding of how PCIDs impact anxiety treatment response, will inform the development of future interventions, and will provide a strong foundation for future work aimed at increasing treatment response rates for anxious youth.